Earth and Planetary Science Letters 286 (2009) 122–130

Contents lists available at ScienceDirect

Earth and Planetary Science Letters

j ourna l homepage: www.e lsev ie r.com/ locate /eps l

Insights into the formation of Fe- and Mg-rich aqueous solutions on early Marsprovided by the ALH 84001 carbonates

Paul B. Niles a,⁎, Mikhail Yu. Zolotov b, Laurie A. Leshin c

a Astromaterials Research Exploration Sciences, Mail Code KR, NASA Johnson Space Center 2101 NASA Parkway, Houston, TX 77058, United Statesb School of Earth and Space Exploration, Arizona State University, Tempe, AZ 85287-1404, United Statesc NASA Goddard Space Flight Center, Greenbelt, MD 20771, United States

⁎ Corresponding author. Tel.: +1 281 483 7860; fax:E-mail address: paul.b.niles@nasa.gov (P.B. Niles).

0012-821X/$ – see front matter © 2009 Published by Edoi:10.1016/j.epsl.2009.06.039

a b s t r a c t

a r t i c l e i n f o

Article history:Received 3 March 2009Received in revised form 15 June 2009Accepted 16 June 2009Available online 19 August 2009

Editor: M.L. Delaney

Keywords:MarsALH 84001carbonatesmeteoriteaqueous solutionscarbon dioxide

The chemical and isotopic pattern of the zoned carbonate globules in the ALH 84001 meteorite reveals aunique aqueous environment on early Mars. If the evolution of the fluid composition was dictated primarilyby carbonate precipitation, the zoning pattern of the carbonates can constrain the fluid to have had an Mg/Camole ratioN~5.3 and a Fe/Ca mole ratioN~1 prior to the formation of the carbonates. Chemical equilibriummodeling of water–rock interactions indicates that low temperatures and low pH favor the formation of anaqueous solution with elevated Mg and Fe concentrations. The modeling shows that a sufficiently Fe- andMg-rich fluid could have formed through low-temperature (b100 °C) subsurface aqueous alteration of anALH 84001-type rock at pH 5–7. This range of pH corresponds to an elevated CO2 fugacity (~0.1–1 bar).Formation of ALH 84001 carbonates could have been driven by degassing of CO2 and corresponding pHincrease in near-surface environments during an upwelling of subsurface CO2-rich solutions. This scenario isconsistent with the unaltered nature of the ALH 84001 rock and with chemical and isotopic composition ofits carbonates.

© 2009 Published by Elsevier B.V.

1. Introduction

The detailed chemical and geological characteristics of aqueoussystems on Mars have long been sought after for informationregarding the history of water. While recent orbital and landedmissions toMars have significantly advanced our understanding of thehistory and abundance of water, this has just exposed many newquestions about how water has interacted with the planet. Resultsfrom the Mars Exploration Rovers (MER) reveal sediments left behindby aqueous systems dominated by acidic sulfate-rich solutions(Squyres et al., 2004; Squyres and Knoll, 2005; Ming et al., 2006).Widespread detection of sulfates on the surface of Mars by the MarsExpress orbiter has prompted some to suggest that they represent awhole epoch where aqueous systems were dominated by sulfur-richacidic solutions and that clay minerals detected in older terrain onMars represent an earlier epoch of more alkaline aqueous environ-ments (Bibring et al., 2006).

Martian meteorites have provided a perspective of aqueoussystems (Bridges et al., 2001) that has been different from the viewprovided by orbital photography and spectroscopy, as well as datafrom landers (e.g. McSween et al., 1999; Bandfield et al., 2000;

+1 281 483 1573.

lsevier B.V.

Craddock and Howard, 2002; Clark et al., 2005). One of the keyreasons for these discrepancies is that many of the Martianmeteoritesaremuch younger than the aqueous events observed, and thus may benot representative of the appropriate ages or locations on Mars toprovide information regarding its aqueous history. This is not true ofALH 84001 whose ancient crystallization age indicates that it hasexperienced almost all of Mars' history. In addition, the 3.9 Ga age ofthe carbonates (Borg et al., 1999) in this specimen places theirformation at the time of late heavy bombardment that could havecaused impact-generated fluvial activity (Segura et al., 2002) andintense erosion by surface fluids (Craddock and Howard, 2002). Theancient age of the ALH 84001 meteorite may indicate that it camefrom the heavily cratered highlands, which preserve valley networksthat have been hypothesized to have formed around the same time asthe ALH 84001 carbonates (Carr and Clow, 1981; Pieri, 1976).

1.1. ALH 84001 carbonates

The carbonates in the ALH 84001 meteorite provide an opportu-nity to understand the details of an ancient aqueous system on Mars.Their unique chemical, isotopic, andmineralogical compositions pointtoward the prospect of making conclusive statements about thegeological conditions inwhich they formed including the compositionof the fluids, temperature, and association with the atmosphere, andpossible influence of life.

123P.B. Niles et al. / Earth and Planetary Science Letters 286 (2009) 122–130

The chemical composition of the ALH 84001 carbonates has beencharacterized using a great variety of analytical techniques includingelectronmicroscopy (Mittlefehldt,1994;McSween,1996; Gilmour et al.,1997; Harvey and Corrigan and Harvey, 2004), time of flight secondaryion mass spectrometry (ToF-SIMS) (Holland et al., 2000), X-rayfluorescence microprobe (Flynn et al., 1997), and Raman spectroscopy(Cooney et al., 1999; Steele et al., 2007). The major element composi-tions are entirely constrained in a single field on the Ca, Mg, and Feternary diagram (Fig. 1). This field ranges from pure calcium carbonateto almost pure magnesium carbonate (Ca5Mg95Fe0) (Harvey andMcSween, 1996; Eiler et al., 2002; Corrigan and Harvey, 2004).

The carbonates appear as zoned globules or fragments thereof thatconsistently grade from more Ca-rich compositions on the interior tomore Mg-rich compositions on their exterior (Fig. 1) (Harvey andMcSween, 1996). It is very uncommon for a single globule to possessthe entire range of compositions, and they frequently only sample themore Mg-rich sequence. However the entire range of chemicalcompositions in a single globule has been observed in rare instances(Corrigan and Harvey, 2004). In other rare instances the sequenceseems to repeat itself with a zoned fragment of Mg-rich carbonateenclosed by more Ca-rich compositions (Ca76Mg12Fe05Mn06) thatgrade into more intermediate compositions (Ca15Mg51Fe32Mn02)(Holland et al., 2005).

Anders (1996) suggested that the chemical variations observed inthe ALH 84001 carbonates are caused by differences in the solubilityproducts of the different carbonate minerals. This notion is consistentwith laboratory experiments of Golden et al. (2001, 2000). Buildingupon the previous studies mentioned above, this work seeks to usethe unusual zoning and Mg-rich chemical compositions of thepredominant ALH 84001 carbonate globules as a constraint for thechemical composition of their original fluid. These constraints canthen be used to understand formation and evolution of these fluidsthrough interactions with rocks and an early Martian atmosphere.

2. Methods

The nature of ALH 84001 as a meteorite precludes any knowledgeof the geologic environment in which it existed when the carbonateswere deposited. For the purposes of this study, we assume that ALH84001 was a typical rock in that geologic environment and that thechemical composition of dissolved species in the water, from whichthe carbonates formed, was primarily influenced by interaction withrocks of a similar composition to ALH 84001.

Fig. 1. The ternary diagram showing the molar composition of ALH 84001 carbonatesindicated by the gray field (Corrigan and Harvey, 2004; Harvey and McSween, 1996).Arrows indicate the proposed evolution of carbonate composition with precipitation. Ingeneral, the earliest forming carbonates are the most Ca-rich, while the later formingcarbonates are the most Mg-rich.

This interaction can be understood using calculations of chemicalequilibria inwater–rock systems. The thermodynamic modeling is notused here to address the carbonate precipitation event; it is insteadaimed at better understanding the fluid–rock interaction thatoccurred prior to the precipitation of the carbonates. Thermodynamicmodeling of fluid–rock interaction has been applied successfully usingthis technique in many studies starting with pioneering works ofHelgeson [e.g. 1979].

A series of equilibrium fluid and secondary mineral compositionswere calculated at various water/rock (W/R) ratios, CO2 fugacities(fCO2), and temperatures at total pressure of 5 bar. Varying the totalpressure did not affect the results in any significant way, and having ahigher pressure allowed for modeling aqueous systems at tempera-tures greater than 100 °C. CO2 fugacity was held at a fixed value, so thesystem was held open to CO2 but remained closed to everything else.The major elements in the bulk composition of ALH 84001 meteorite(Lodders,1998) were included in our system (H–O–C–Fe–Mg–Ca–Na–Al–Si) and the ALH 84001 bulk composition was normalized toexclude theminor elements. The simplified composition of ALH 84001type rock was as follows (inwt.%): SiO2: 53.58, Na2O: 0.14, Al2O3: 1.31,FeO: 17.76, CaO: 1.85, andMgO: 25.37.We used the GEOCHEQ program(Mironenko et al., 2008) to calculate chemical equilibria using the freeenergy minimization method. The minerals, gases and aqueousspecies included in the geochemical modeling are shown in Table 1.The redox potential of the system was obtained through equilibriumcalculations.

Kinetic factors can play a significant role in inhibiting mineraldissolution and precipitation at the low temperatures (25 °C) used inthis study(Burns, 1993; Pfeifer, 1977; Brantley et al., 2003). We canaccount for some of these effects by suppressing certain minerals in ourcalculations which prevents them from forming. However, the calcula-tions do assume stoichiometric dissolution of the primary mineralphases and this is can be complicated by kinetic factors (Burns,1993). Inthis case, ALH 84001 ismade up almost entirely of orthopyroxenewhichshows stoichiometric dissolution over longer timescales under moder-ately acidic conditions (Burns, 1993; Brantley and Chen, 1995).

At temperatures 50 °C and below, we excluded formation ofantigorite, talc, tremolite, Mg anthophyllite, magnesite, huntite,dolomite, siderite, and graphite. These exclusions were based on thecomposition of solutions associated with ultramafic rocks on Earth(Bruni et al., 2002; Pfeifer, 1977). Most of these fluids show strongsupersaturation for the above minerals below ~50 °C. For example,dolomite andmagnesite are supersaturated in thesewaters as they arekinetically inhibited from precipitating at lower temperatures(Lippmann, 1973). Hydromagnesite and calcite/aragonite do notshow these supersaturations and are also frequently observed inlow-temperature alteration assemblages (e.g. O'Neil and Barnes,1971). Pfeifer (1977) indicated that sepiolite was among this set ofminerals that did not precipitate at low temperatures, however,sepiolite is found in low-temperature alteration assemblages ofultramafic rocks (Holland, 1978; Bruni et al., 2002). Therefore,sepiolite was included in our low-temperature calculations. Attemperatures below 100 °C antigorite, talc, graphite, and methaneare excluded from the calculations. Methane was excluded because ofits very slow abiotic formation at low temperatures (McCollom andSeewald, 2007).

The water/rock mass ratios used vary from 1 to 104. The very highW/R ratios are useful for approximating the scenario where the waterhas reacted with a small amount of rock, for example, at the beginningof alteration. In this way, decreasing W/R ratio can be used as a proxyfor rock alteration progress (Helgeson, 1979).

3. Results

We calculated equilibrium compositions in water–rock system at avariety of water/rock ratios, temperatures, and fCO2 to evaluate the

Fig. 2. The calculated concentrations of aqueous species versus W/R ratio at 25 °C andan fCO2 of 0.1 bar. At the highest W/R ratios, the concentration of aqueous speciesresembles the concentration of the host rock as secondary phases do not form. Aswater/rock ratios decrease below 100, there is very little effect on the relativeproportions of Mg2+, Fe2+, and Ca2+.

Table 1Minerals used in equilibrium thermodynamic modeling. Thermodynamic data fromShock et al. (1989), Shock et al. (1997), and Helgeson et al. (1978).

MineralAkermanite SepioliteAmorphous Silica SideriteAndradite TalcAnorthite TremoliteAnthophyllite WairakiteAntigorite WollastoniteAragonite ZoisiteArtinite Aqueous SpeciesBoehmite Al3+

Brucite AlO+

Ca–Al–Pyroxene AlO2−

Calcite AlOH2+

Chalcedony Ca2+

Chrysotile CaCO3,aqClinochlore CaHCO3

+

Clinozoisite CaHSiO3+

Goethite CaOH+

Cordierite, hydrous CO,aqDaphnite,14A CO2,aqDiaspore CO3

2−

Diopside Fe2+

Dolomite Fe3+

Enstatite FeO,aqEpidote FeO+

Fayalite FeO2−

Ferrosilite FeOH+

Ferrotremolite FeOH2+

Forsterite H+

Gibbsite H2,aqGraphite HAlO2,aqGreenalite HCO3

−

Grossular HFeO2−

Hedenbergite HFeO2,aqHematite HSiO3

−

Huntite Mg2+

Hydromagnesite MgCl+

Kaolinite MgCO3,aqLaumontite MgHCO3+

Lawsonite MgHSiO3+

Magnesite MgOH+

Magnetite O2,aqMargarite OH−

Monticellite GasesNesquehonite COPortlandite CO2

Prehnite H2

Pyrophyllite H2OQuartz O2

124 P.B. Niles et al. / Earth and Planetary Science Letters 286 (2009) 122–130

effect of each variable on solution composition and secondarymineralogy. The calculations show that the solution compositiondoes not vary significantly at W/R between 1 and 100 (Fig. 2). At W/Rb1000, solution composition reflects precipitation of mineralphases. Calculated solutions are very dilute at the highest W/R ratios(N103), precipitated minerals were not abundant, and did notsignificantly affect solution composition. At these high W/R ratios,the solution composition is similar to the overall bulk composition ofthe rock.

Over all of the considered W/R ratios, temperatures below ~100 °Ctend to yield solutions with higher Mg/Ca and Fe/Ca ratios whilehigher temperatures yielded solutions withmuch lower ratios (Fig. 3).The concentrations of Fe2+ and Mg2+ are strongly affected bytemperature which acts to decrease the solubilities of their parentminerals (Fig. 4). The concentration of Ca2+ is not strongly affected byincreasing temperatures and thus Mg/Ca and Fe/Ca ratios decreasesteadily with increasing temperature. The concentration of Fe2+ isdepressed by the formation of Fe-rich serpentine (greenalite) attemperatures N100 °C. Likewise, Mg2+ concentrations decrease withthe precipitation of Mg-rich serpentine at higher temperatures. The

concentration of Ca2+ is not affected as it precipitates out only in Ca-bearing carbonate which does not increase significantly at highertemperatures (Fig. 5). Under most conditions used in this study, Ca-bearing minerals were not stable at higher temperatures while Mgand Fe bearing minerals were dominant. At temperatures 50–100 °C,secondary mineralogy is typically dominated by carbonate mineralssuch as siderite, magnesite, and dolomite. Under the conditions oflower temperatures (b50 °C) and higher fCO2, pH decreases (Fig. 6)and secondary minerals include silica, calcite, hydromagnesite, Fe-,Mg-rich chlorite, Fe-, Mg-rich serpentine minerals, and kaolinite.

Fugacity of CO2 also strongly influences the concentrations of Mg2+

and Fe2+ in solution (Fig. 6). Increasing fCO2 results in a decrease in pH,allowing for increased solubility of Mg- and Fe-rich minerals (Fig. 6). ThiscausesMg/Ca and Fe/Ca ratios to increasewith increasing fCO2 since Ca2+

remains less affected. The fugacityof CO2 is theprimarydriverof pH in thissystem (Fig. 6), with higher fCO2 resulting in lower pHwhich ranges from4.7 to 9.9 throughout the range of conditions considered. At anytemperature, fluids in equilibrium with fCO2 below 0.1 bar are depletedin Fe2+ and Mg2+ relative to Ca2+ (Fig. 8).

4. Discussion

4.1. Constraints on original fluid composition

Strong constraints can be placed on the chemical composition ofthe fluid that formed the ALH 84001 carbonates if a few assumptionsare made. The first assumption is that the Ca-rich carbonatesprecipitated first followed by the Mg-rich carbonates. This is stronglysupported by the petrographic relationships documented in a numberof studies (Harvey and McSween, 1996) where Ca-rich phases arelocated on the interior of the carbonate globules and Mg-richcompositions are located on the outer edges.

The second assumption is that the ALH 84001 carbonate globulesformed from a single fluid. This is supported by the following: (1) thechemical composition of the globules varies smoothly and evenly fromcore to rim according to differences in solubilities of each carbonatecomposition (Anders, 1996); (2) the oxygen isotope composition ofthe carbonates varies smoothly with Mg-content (Leshin et al., 1998)implying the evolution of a single system rather than fluid mixing;(3) laboratory experiments using a single fluid can successfullyrecreate the Mg-rich rims observed on the ALH 84001 carbonates(Golden et al., 2000); and (4) the petrographic occurrence of globulesas single continuous entities that suggest formation in a single event(Treiman, 1995).

Fig. 3. The calculated Mg/Ca and Fe/Ca mole ratios in aqueous solution for different W/R ratios and temperatures. Colors indicate either Mg/Ca ratio or Fe/Ca ratio. In each case,warmer colors represent solution compositions that are consistent with the constraints provided by the ALH 84001 carbonates while dark purple indicates solution compositionsincompatible with our constraints. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

125P.B. Niles et al. / Earth and Planetary Science Letters 286 (2009) 122–130

The third assumption is that the chemical composition of the fluidevolved solely due to the precipitation of carbonates that were moreCa-rich than an original bulk composition of aqueous solution. Theminimum initial Mg/Ca ratio of the formation fluid is dictated by thebulk chemical composition of the carbonates and the Mg-rich natureof the final phases. If the composition of the fluid changed through theprecipitation of the carbonates, a significant portion of the Ca2+ in thefluid must have been consumed by carbonate precipitation in order tosignificantly change the Mg/Ca ratio in the fluid. It follows that if theinitial Mg/Ca ratio of the fluid was less than the average Mg/Ca ratioof the carbonate globules, then the solution would have a lower Mg/Ca ratio after precipitating the carbonate. This is inconsistent with thechemical composition of the carbonates, which show an ever-increasing Mg/Ca ratio as precipitation proceeded. Thus, the Mg/Caratio of the fluid must have remained greater than the Mg/Ca ratio of

the carbonate throughout the precipitation of the globules. Thus, theinitial fluid must have had an Mg/Ca ratio that was greater than theMg/Ca ratio of the bulk carbonates, which is ~5.3 (Mittlefehldt, 1994).Likewise, the high Fe content of the carbonates indicates high ironcontent in the original solution.

The initial Mg/Ca ratio of the solution is an important factoraffecting the chemical composition of the precipitated carbonates. Ifthe initial Mg/Ca ratio was too high, the fluid would have beenincapable of precipitating Ca carbonate during the initial stages of thereaction. The precipitation of Ca carbonates has been observed inlaboratory experiments with Mg/Ca ratios as high as 30 at 20 °C(Pokrovsky, 1996). The preference for Ca carbonates over Mgcarbonates even at high Mg/Ca ratios is due to kinetic effectsinhibiting Mg-rich carbonate formation associated with the behaviorof the Mg2+ ion (Lippmann, 1973). As the initial fluid temperature

Fig. 4. The concentrations of aqueous species versus temperature at W/R ratio of 100and an fCO2 of 0.5 bar. Above 100 °C Fe/Ca ratios plunge well below 0.1, an order ofmagnitude lower than the constraint poised by the ALH 84001 carbonates.

126 P.B. Niles et al. / Earth and Planetary Science Letters 286 (2009) 122–130

increases, the maximum Mg/Ca ratio that still allows Ca carbonateprecipitation decreases (Pokrovsky, 1996).

The Fe/Ca ratio of the solution was also elevated because the ALH84001 carbonates incorporate significant amounts of iron. The highaverage Fe/Ca ratio in the carbonates (~1) dictates that the Fe2+

concentration of the solution needed to be comparable to the Ca2+

concentration. The last carbonates to precipitate are Mg-rich and are,

Fig. 5. Calculated molar amounts of cations incorporated into secondary minerals atdifferent temperatures for 10 g rock,W/R=100 and fCO2 of 0.5 bar. The y-axis indicatesthe number of moles of either Ca, Mg or Fe depending on the mineral.

Fig. 6. Top plot shows concentrations of aqueous solutes versus fCO2 atW/R=102 and atemperature of 25 °C in a water–rock system where rock is presented by ALH 84001-type composition. Bottom plot shows the pH of aqueous solution versus fCO2 attemperature of 25 °C, reacted with an ALH 84001-type rock at different W/R ratios.

therefore, depleted of both Fe and Ca. This suggests that both Fe and Cabecame significantly depleted through the initial precipitation of thecarbonates. While it is possible that there was much more Fe2+ in thesolution initially and that it was lowered through the precipitation ofminerals, it is very unlikely that the initial Fe2+ concentration wassignificantly lower than the Ca2+ concentration because of the Fe-richnature of the carbonate globules. Thus, an initial solution with a Fe/Caratio that was less than ~1 by an order of magnitude or more wouldnot be capable of precipitating the ALH 84001 carbonates.

4.2. Insights from modeling results

The results from the thermodynamicmodeling show that solutionswith high concentrations of Mg2+ and Fe2+ could form through aninteraction with the ALH 84001 type rocks at lower temperatures andhigher fCO2. Despite the Mg-, and Fe-rich nature of the rock, suitablesolutions for Mg/Ca and Fe/Ca ratios are only produced at lowtemperatures (b100 °C) and high fCO2 (partial pressuresN0.1 bar), orW/RN104 (Figs. 3 and 8). Outside of these limits, Fe/Ca ratios becomelower than the constraint of ~1 by more than an order of magnitude(Fig. 4). These limits are dictated by the secondary minerals formedunder those conditions. At temperatures N100 °C, Fe2+ and Mg2+

precipitate into serpentine minerals and iron oxides while Ca2+

remains in the solution (Figs. 4 and 5). The partial pressure of CO2

strongly affects the pH of the solution and higher pH enhances thestability of serpentine minerals (Fig. 7). Water/rock ratios above ~104

were not considered because they would not produce solutionssaturated with respect to carbonates.

Fig. 7. Moles of cations incorporated into secondary minerals versus fCO2 at 25 °C andwith 10 g of rock and W/R=100. Both Fe- and Mg-rich silicates have greater solubilitywith increasing fCO2 (and corresponding decreasing pH), while Ca-bearing carbonatesare not strongly affected in the pH ranges modeled. The y-axis indicates the number ofmoles of either Fe, Ca, or Mg depending on the mineral. Goethite is present at all CO2

fugacities but only in very small amounts that are not visible on the graph.

127P.B. Niles et al. / Earth and Planetary Science Letters 286 (2009) 122–130

While we rule out very dilute solutions formed from water/rockratios greater than ~104; solutions formed with low W/R (b~102) arealso unlikely. Equilibrium compositions calculated at low water/rockratios may represent extensive aqueous alteration of the host rock.Such an alteration is inconsistent with the oxygen isotope composi-tion of the ALH 84001 carbonates that indicate that they are not inequilibrium with the host rock (Farquhar et al., 1998; Treiman andRomanek, 1998) and that significant water–rock interaction did nothappen. Thus, a low temperature alteration (at 100bW/Rb10000) ofa rock similar to ALH 84001 by solutions with moderate pH (possiblycaused by elevated CO2) is consistent with these isotopic data. This isalso consistent with the composition of spring waters produced fromultramafic rocks (Barnes and O'Neil, 1969; Pfeifer, 1977; Barnes et al.,1982; Bruni et al., 2002). Studies of the evolution and development ofwaters reacting with ultramafic rocks have shown that the inter-mediate pH (6–9) magnesium bicarbonate rich waters are the resultof limited interaction with the ultramafic rocks, while fluids withpHN~10 are the product of extensive interactions (Bruni et al., 2002).

The presence of Fe-rich carbonate in ALH 84001 without Fe-richclayminerals similar to “iddingsite” found in othermartianmeteorites(Bridges et al., 2001) is inconsistent with strongly alkaline pH andmayimply CO2-rich (fCO2N0.1 bar) environment of carbonate deposition.The siderite-Fe2+-phyllosilicates transition at fCO2=0.1 and 25 °C(Catling, 1999) is similar to the 0.1 bar threshold observed in thisstudy. A CO2-rich solution is also consistent with theMg-rich nature ofthe ALH 84001 carbonates, since Mg-rich carbonates are favored insystems with elevated CO2 concentrations (Pokrovsky, 1996).

Other species such as sulfur oxides, HCl, and oxidizing sulfideshave been proposed to be responsible for very acidic (pHb3) solutionson Mars (Banin et al., 1997; Zolotov and Shock, 2005). This is probablynot the case for the solution that formed the ALH 84001 carbonates.

The presence of carbonates suggests that the solutionwas atmost onlymildly acidic, and the lack of sulfates and chlorides in ALH 84001suggests that S and Cl were not dominant species in the solutionwhenthe carbonates precipitated. Therefore we favor CO2 to be the cause ofthe mildly acidic conditions (pH=5–7). Degassing of CO2 could alsobe responsible for the carbonate deposition, as discussed below.

4.2.1. Deposition conditions of ALH 84001 carbonatesTwo dominant views have emerged regarding the formation

environment of the ALH 84001 carbonates. One hypothesis contendsthat the ALH 84001 carbonates have a number of characteristics thatindicate their low temperature (b~100 °C) deposition in a dynamicenvironment (Valley et al., 1997; Warren, 1998; McSween and Harvey,1998; Niles et al., 2005). While the other hypothesis suggests that thecarbonates formed in a rapidly cooling environment which is initiallyhigher temperature (N150 °C) and cools to ~30 °C (Romanek et al.,1994; Eiler et al., 2002; Steele et al., 2007). Both of these views attemptto tie together the wide variety of data collected from this meteoriteand synthesize it into a reasonable scenario.

The results of this study suggest that the ALH 84001 carbonatesformed from an Mg-, Fe-rich fluid whose temperature did not exceed~100 °C (Fig. 3) and was initially moderately acidic (pH=5–7). Ourwork demonstrates that acidic pH could be due to elevated fCO2. Thisseems to support the low temperature hypothesis because at highertemperatures Fe2+ would be precipitated. However, as total pressureincreases above 500 bars higher Fe2+ concentrations can be obtainedin the fluid at higher temperatures (Koziol, 2004). This could allowhigher temperature fluids (N100 °C) to be capable of precipitating theALH 84001 globules, but much higher total pressures than those usedin this study (5 bars) must be present to retain enough Fe2+ in thesolution.

The results of this study further constrain the conditions of a lowtemperature formation environment to have a relatively low pH (5–7).Thus higher pH fluids like those proposed in Niles et al. (2005)wouldnot be sufficiently Mg- or Fe-rich in order to precipitate carbonateslike the globules in ALH 84001. However, the results of the currentstudy provide a different pathway for reconciling the chemical,isotopic, andmineralogic compositions of the ALH 84001 carbonates.Carbon dioxide rich fluids, similar to those modeled in this study, cancreate large carbon isotope enrichments in carbonates through rapidCO2 degassing caused by freezing, evaporation, and/or boiling, whichcan be caused by depressurization of fluids at near-surface conditions(Michaelis et al., 1985; Zheng, 1990; Clark and Lauriol, 1992; Mickleret al., 2006).

We propose that the ALH 84001 carbonates could have formedduring a subsurface fluid mobilization event where CO2-rich fluidswere exposed to a low-pressure environment similar to the atmo-sphere of Mars today. Carbon dioxide rich subsurface solutions wouldundergo rapid depressurization upon reaching the surface causingCO2 degassing through bubbling and evaporation, and carbonateprecipitation, which can be represented by the well-known reaction:

2HCO−3 þ Ca

2þ→CaCO3 þ H2O þ CO2:

Previously, McSween and Harvey (1998) and Warren (1998) hadproposed that evaporation could be a mechanism for creating thecarbonates. This is inconsistent with the δ18O compositions of thecarbonates if we consider the retrograde behavior of δ18O enrichment inevaporitic environments on Earth (Sofer and Gat, 1975). However, if theclimate of early Mars was similar to the extremely dry currentconditions, too little water vapor would be present in the atmosphereto cause retrograde δ18O enrichments (Lloyd, 1966; Stiller et al., 1985;Yadav, 1997). Therefore, we might expect that rapid H2O evaporation(including boilingwell below100 °C) into dry low-pressure atmospherecoupledwith abundant CO2 degassing to be capable of creating δ18O andδ13C rich solutions.

Fig. 8. The Mg/Ca and Fe/Ca mole ratios as functions of the W/R ratio and fCO2. Colors indicate either Mg/Ca ratio or Fe/Ca ratio. In each case, warmer colors represent solutioncompositions that are consistent with the constraints provided by the ALH 84001 carbonates while dark purple indicates solution compositions incompatible with our constraints.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

128 P.B. Niles et al. / Earth and Planetary Science Letters 286 (2009) 122–130

This hypothesis requires three important elements: 1) Themartianatmosphere was low-pressure and dry 3.9 Ga ago, 2) CO2-rich waterswere formed in the subsurface in spite of the sparse atmosphere, and3) these fluids moved upward and approached near-surface perme-able materials, and could have released upon the surface in a spring-type environment.

Current climatic models indicate that ancient Mars may have had alow CO2-pressure atmosphere similar to today's (Haberle, 1998). Inaddition, the mass independent fractionation of oxygen isotopesobserved in ALH 84001 carbonates is inconsistent with a denseatmosphere and active hydrosphere at that time (Farquhar et al.,1998). Subsurface CO2-rich waters could have formed through degas-sing of magma chambers, which also favored local warming of ice-bearing rocks and development of circulation of crustal fluids. Anothersource of CO2 would be high-temperature decomposition of earlier-formed crustal carbonates caused by magmatic or impact events (Carr,1989). In addition, a mixture of water–ice and CO2-clathrates (Dobro-volskis and Ingersoll, 1975; Longhi, 2006) in the ancient subsurface ofMars could form CO2-rich waters in the absence of a CO2-richatmosphere. The surface manifestation of fluid upwelling might lookvery similar to the valley networks, many of which are located on theancient terrain of the southern highlands (Pieri, 1976).

5. Conclusions

Given three simple assumptions that the Ca-rich carbonates in ALH84001 formed first followed by the Mg-rich carbonates, that thecarbonate rosettes formed from a single fluid, and that their chemicalvariation was driven by carbonate precipitation; the chemicalcomposition of the carbonates dictates that their initial formation

fluid must have had anMg/Ca ratio greater than ~5.3 and a Fe/Ca ratiogreater than ~1.

Equilibrium thermodynamic calculations shows that Fe- and Mg-richsolutions can be produced onMars through low-temperature (b~100 °C)aqueous alteration of ALH 84001 like ultramafic rocks at fCO2N0.1 bar.This suggests that CO2-rich solutions may play an important role informing Fe- and Mg-rich carbonates on Mars. These solutions also havechemical compositions which are consistent with the formation of theALH 84001 carbonates.

Equilibrium thermodynamic modeling indicates that fluids inequilibrium with ALH 84001 type rocks at temperatures greater than100 °C would not be capable of precipitating the ALH 84001 carbonatesdue to the formation of abundant iron oxides and serpentine minerals.The modeling also indicates that fluids in equilibrium with ALH 84001type rocks at neutral to alkaline pH would not be able to maintain highenough Mg/Ca and Fe/Ca ratios in the solution. Thus, some means formaintaining a lower pH (b~7) (possibly high due to elevated fCO2) isnecessary to prevent loss of Mg2+ and Fe2+ from the solution throughprecipitation of serpentine and iron oxides.

Exposure of CO2-rich subsurface fluids to a low pressure earlymartian near-surface could create an environment that is consistentwith the formation of the chemical, isotopic and mineralogiccharacteristics of the ALH 84001 carbonates through the evaporationof water from the solution and CO2 degassing.

Acknowledgements

Critical reviews and recommendations made by Nick Tosca and ananonymous reviewer improved this manuscript and are greatlyappreciated. This material is based upon the work supported by grants

129P.B. Niles et al. / Earth and Planetary Science Letters 286 (2009) 122–130

from theNational Aeronautics and Space Administration issued throughtheMars Fundamental ResearchProgramtoDr.Niles andDr. Zolotov.Wethank Mikhail Mironenko for sharing the GEOCHEQ code which wasused for these calculations.

References

Anders, E., 1996. Evaluating the evidence for past life on Mars. Science 274, 2119–2121.Bandfield, J.L., Hamilton, V.E., Christensen, P.R., 2000. A global view of Martian surface

compositions from MGS-TES. Science 287, 1626–1630.Banin, A., Han, F.X., Kan, I., Cicelsky, A., 1997. Acidic volatiles and the Mars soil.

J. Geophys. Res.-Planets 102, 13341–13356.Barnes, I., O'Neil, J.R., 1969. Relationship between fluids in some fresh alpine-type

ultramafics and possible modern serpentinization, Western United States. Geol.Soc. Am. Bull. 80, 1947–1960.

Barnes, I., Presser, T.S., Saines, M., Dickson, P., Vangroos, A.F.K., 1982. Geochemistry ofhighly basic calcium hydroxide groundwater in Jordan. Chem. Geol. 35, 147–154.

Bibring, J.P., Langevin, Y., Mustard, J.F., Poulet, F., Arvidson, R., Gendrin, A., Gondet, B.,Mangold,N., Pinet, P., Forget, F., the, O.t., Berthe,M., Bibring, J.-P., Gendrin, A., Gomez, C.,Gondet, B., Jouglet, D., Poulet, F., Soufflot, A., Vincendon, M., Combes, M., Drossart, P.,Encrenaz, T., Fouchet, T., Merchiorri, R., Belluci, G., Altieri, F., Formisano, V.,Capaccioni, F., Cerroni, P., Coradini, A., Fonti, S., Korablev, O., Kottsov, V., Ignatiev, N.,Moroz, V., Titov, D., Zasova, L., Loiseau, D., Mangold, N., Pinet, P., Doute, S., Schmitt, B.,Sotin, C., Hauber, E., Hoffmann, H., Jaumann, R., Keller, U., Arvidson, R., Mustard, J.F.,Duxbury, T., Forget, F., Neukum, G., 2006. Global mineralogical and aqueous Marshistory derived from OMEGA/Mars Express data. Science 312, 400–404.

Borg, L.E., Connelly, J.N., Nyquist, L.E., Shih, C.Y., Wiesmann, H., Reese, Y., 1999. The age ofthe carbonates in Martian meteorite ALH84001. Science 286, 90–94.

Brantley, S.L., Chen, Y., 1995. Chemical weathering rates of pyroxenes and amphiboles.Chemical Weathering Rates of Silicate Minerals. Reviews in Mineralogy, vol. 31.Mineralogical Soc America, Washington, pp. 119–172.

Brantley, S.L., Heinrich, D.H., Karl, K.T., 2003. Reaction Kinetics of Primary Rock-FormingMinerals under Ambient Conditions. Treatise on Geochemistry. Pergamon, Oxford,pp. 73–117.

Bridges, J.C., Catling, D.C., Saxton, J.M., Swindle, T.D., Lyon, I.C., Grady, M.M., 2001.Alteration assemblages in Martian meteorites: implications for near-surfaceprocesses. Space Sci. Rev. 96, 365–392.

Bruni, J., Canepa, M., Chiodini, G., Cioni, R., Cipolli, F., Longinelli, A., Marini, L., Ottonello,G., Zuccolini, M.V., 2002. Irreversible water–rock mass transfer accompanying thegeneration of the neutral, Mg-HCO3 and high-pH, Ca-OH spring waters of theGenova province, Italy. Appl. Geochem. 17, 455–474.

Burns, R.G., 1993. Rates and mechanisms of chemical-weathering of ferromagnesiansilicate minerals on Mars. Geochim. Cosmochim. Acta 57, 4555–4574.

Carr, M.H., 1989. Recharge of the early atmosphere of Mars by impact-induced release ofCO2. Icarus 79, 311–327.

Carr, M.H., Clow, G.D., 1981. Martian channels and valleys — their characteristics,distribution, and age. Icarus 48, 91–117.

Catling, D.C., 1999. A chemical model for evaporites on earlyMars: possible sedimentarytracers of the early climate and implications for exploration. J. Geophys. Res.-Planets 104, 16453–16469.

Clark, B.C., Morris, R.V., McLennan, S.M., Gellert, R., Jolliff, B., Knoll, A.H., Squyres, S.W.,Lowenstein, T.K., Ming, D.W., Tosca, N.J., 2005. Chemistry and mineralogy ofoutcrops at Meridiani Planum. Earth Planet. Sci. Lett. 240, 73–94.

Clark, I.D., Lauriol, B., 1992. Kinetic enrichment of stable isotopes in cryogenic calcites.Chem. Geol. 102, 217–228.

Cooney, T.F., Scott, E.R.D., Krot, A.N., Sharma, S.K., Yamaguchi, A., 1999. Vibrationalspectroscopic study of minerals in the Martian meteorite ALH84001. Am. Mineral.84, 1569–1576.

Corrigan, C.M., Harvey, R.P., 2004. Multi-generational carbonate assemblages in Martianmeteorite Allan Hills 84001: implications for nucleation, growth, and alteration.Meteorit. Planet. Sci. 39, 17–30.

Craddock, R.A., Howard, A.D., 2002. The case for rainfall on a warm, wet early Mars.J. Geophys. Res.-Planets 107.

Dobrovolskis, A., Ingersoll, A.P., 1975. Carbon dioxide-water clathrate as a reservoir ofCO2 on Mars. Icarus 26, 353–357.

Eiler, J.M., Valley, J.W., Graham, C.M., Fournelle, J., 2002. Two populations of carbonate inALH84001: geochemical evidence for discrimination and genesis. Geochim.Cosmochim. Acta 66, 1285–1303.

Farquhar, J., Thiemens,M.H., Jackson, T.,1998. Atmosphere–surface interactions onMars:delta O-17 measurements of carbonate from ALH 84001. Science 280, 1580–1582.

Flynn, G.J., Sutton, S.R., Keller, L.P., 1997. Element abundance patterns in carbonateglobules and rims from ALH 84001. Lunar Planet. Sci. Conf. XXVIII, Abstract #367.

Gilmour, J.D., Wogelius, R.A., Grime, G.W., Turner, G., 1997. Trace-element and major-element zoning in Allan Hills 84001 carbonate. Meteorit. Planet. Sci. 32, A48–A49.

Golden, D.C., Ming, D.W., Schwandt, C.S., Lauer, H.V., Socki, R.A., Morris, R.V., Lofgren, G.E.,McKay, G.A., 2001. A simple inorganic process for formation of carbonates, magnetite,and sulfides in Martian meteorite ALH84001. Am. Mineral. 86, 370–375.

Golden, D.C., Ming, D.W., Schwandt, C.S., Morris, R.V., Yang, S.V., Lofgren, G.E., 2000. Anexperimental study on kinetically-driven precipitation of calcium–magnesium–ironcarbonates from solution: implications for the low-temperature formation ofcarbonates in Martian meteorite Allan Hills 84001. Meteorit. Planet. Sci. 35, 457–465.

Haberle, R.M.,1998. EarlyMars climatemodels. J. Geophys. Res.-Planets 103, 28467–28479.Harvey, R.P., McSween, H.Y.,1996. A possible high-temperature origin for the carbonates

in the Martian meteorite ALH84001. Nature 382, 49–51.

Helgeson, H.C., 1979. Mass transfer among minerals and hydrothermal solutions, In:Barnes, H.L. (Ed.), Geochemistry of Hydrothermal Ore Deposits, 2nd ed. JohnWiley,New York, pp. 568–610.

Helgeson, H.C., Delany, J.M., Nesbitt, H.W., Bird, D.K., 1978. Summary and critique of thethermodynamic properties of rock-forming minerals. Am. J. Sci. 278, 1–229.

Holland, G., Lyon, I.C., Cliff, B., Lockyer, N.P., Vickerman, J.C., 2000. Halogen concentrationand distribution in Allan Hills 84001 measured by time-of-flight secondary ion massspectrometry. Meteorit. Planet. Sci. 35, A76-A76.

Holland, G., Saxton, J.M., Lyon, I.C., Turner, G., 2005. Negative delta O-18 values in AllanHills 84001 carbonate: possible evidence for water precipitation onMars. Geochim.Cosmochim. Acta 69, 1359–1370.

Holland, H.D., 1978. The chemistry of the atmosphere and oceans. Wiley, New York.Koziol, A.M., 2004. Experimental determination of siderite stability and application to

Martian meteorite ALH84001. Am. Mineral. 89, 294–300.Leshin, L.A., McKeegan, K.D., Carpenter, P.K., Harvey, R.P., 1998. Oxygen isotopic

constraints on the genesis of carbonates from Martian meteorite ALH84001.Geochim. Cosmochim. Acta 62, 3–13.

Lippmann, F., 1973. Sedimentary carbonate minerals. Springer-Verlag, Berlin-Heidelberg,Federal Republic of Germany.

Lloyd, R.M., 1966. Oxygen isotope enrichment of sea water by evaporation. Geochim.Cosmochim. Acta 30, 801–814.

Lodders, K., 1998. A Survey of SNC Meteorite Whole-Rock Compositions. Meteoritics &Planetary Science, vol. 33, Supplement, p.183–190 33, 183–190.

Longhi, J., 2006. Phase equilibrium in the systemCO2–H2OApplication toMars. J. Geophys.Res.-Planets 111.

McCollom, T.M., Seewald, J.S., 2007. Abiotic synthesis of organic compounds in deep-seahydrothermal environments. Chem. Rev. 107, 382–401.

McSween, H.Y., Harvey, R.P., 1998. An evaporation model for formation of carbonates inthe ALH84001 Martian meteorite. Int. Geol. Rev. 40, 774–783.

McSween, H.Y., Murchie, S.L., Crisp, J.A., Bridges, N.T., Anderson, R.C., Bell, J.F., Britt, D.T.,Bruckner, J., Dreibus, G., Economou, T., Ghosh, A., Golombek, M.P., Greenwood, J.P.,Johnson, J.R., Moore, H.J., Morris, R.V., Parker, T.J., Rieder, R., Singer, R., Wanke, H.,1999. Chemical, multispectral, and textural constraints on the composition andorigin of rocks at the Mars Pathfinder landing site. J. Geophys. Res.-Planets 104,8679–8715.

Michaelis, J., Usdowski, E., Menschel, G., 1985. Partitioning of C-13 and C-12 on thedegassing of CO2 and the precipitation of calcite – Rayleigh-type fractionation and akinetic-model. Am. J. Sci. 285, 318–327.

Mickler, P.J., Stern, L.A., Banner, J.L., 2006. Large kinetic isotope effects in modernspeleothems. Geol. Soc. Am. Bull. 118, 65–81.

Ming, D.W., Mittlefehldt, D.W., Morris, R.V., Golden, D.C., Gellert, R., Yen, A., Clark, B.C.,Squyres, S.W., Farrand,W.H., Ruff, S.W., Arvidson, R.E., Klingelhofer, G., McSween, H.Y.,Rodionov, D.S., Schroder, C., de Souza, P.A., Wang, A., 2006. Geochemical andmineralogical indicators for aqueous processes in the Columbia Hills of Gusev crater.Mars. J. Geophys. Res.-Planets 111.

Mironenko, M.V., Melikhova, T.Y., Zolotov, M.Y., Akinfiev, N.N., 2008. GEOCHEQ_M:Program complex for thermodynamic and kinetic modeling of geochemicalprocesses in rock–water–gas systems. Version 2008. Herald DGGGMS RAS 1.

Mittlefehldt, D.W.,1994. ALH84001, a cumulate orthopyroxenitemember of theMartianmeteorite clan. Meteoritics 29, 214–221.

Niles, P.B., Leshin, L.A., Guan, Y., 2005. Microscale carbon isotope variability inALH84001 carbonates and a discussion of possible formation environments.Geochim. Cosmochim. Acta 69, 2931–2944.

O'Neil, J.R., Barnes, I., 1971. C13 and O18 compositions in some fresh-water carbonatesassociated with ultramafic rocks and serpentinites — Western United-States.Geochim. Cosmochim. Acta 35, 687–697.

Pfeifer, H.R.,1977. Amodel for fluids inmetamorphosed ultramafic rocks. Observation atsurface and subsurface conditions (high pH spring waters). Schweiz. Mineral.Petrogr. Mitt. 57, 361–396.

Pieri, D., 1976. Distribution of small channels on Martian surface. Icarus 27, 25–50.Pokrovsky, O.S., 1996. Experimental synthesis of calcium and magnesium carbonates:

factors controlling magnesian calcite formation in natural water. Lithol. Miner.Resour. 31, 474–482.

Romanek, C.S., Grady, M.M., Wright, I.P., Mittlefehldt, D.W., Socki, R.A., Pillinger, C.T.,Gibson, E.K., 1994. Record of fluid–rock interactions on Mars from the meteoriteALH84001. Nature 372, 655–657.

Segura, T.L., Toon, O.B., Colaprete, A., Zahnle, K., 2002. Environmental effects of largeimpacts on Mars. Science 298, 1977–1980.

Shock, E.L., Helgeson, H.C., Sverjensky, D.A., 1989. Calculation of the thermodynamicand transport-properties of aqueous species at high-pressures and temperatures —standard partial molal properties of inorganic neutral species. Geochim. Cosmo-chim. Acta 53, 2157–2183.

Shock, E.L., Sassani, D.C., Willis, M., Sverjensky, D.A., 1997. Inorganic species in geologicfluids: correlations among standard molal thermodynamic properties of aqueousions and hydroxide complexes. Geochim. Cosmochim. Acta 61, 907–950.

Sofer, Z., Gat, J.R., 1975. Isotope composition of evaporating brines — effect of isotopicactivity ratio in saline solutions. Earth Planet. Sci. Lett. 26, 179–186.

Squyres, S.W.,Grotzinger, J.P.,Arvidson, R.E., Bell, J.F., Calvin,W.,Christensen, P.R., Clark, B.C.,Crisp, J.A., Farrand, W.H., Herkenhoff, K.E., Johnson, J.R., Klingelhofer, G., Knoll, A.H.,McLennan, S.M., McSween, H.Y., Morris, R.V., Rice, J.W., Rieder, R., Soderblom, L.A.,2004. In situ evidence for anancient aqueous environment atMeridiani Planum,Mars.Science 306, 1709–1714.

Squyres, S.W., Knoll, A.H., 2005. Sedimentary rocks at Meridiani Planum: origin,diagenesis, and implications for life on Mars. Earth Planet. Sci. Lett. 240, 1–10.

Steele, A., Fries, M.D., Amundsen, H.E.F., Mysen, B.O., Fogel, M.L., Schweizer,M., Boctor, N.Z.,2007. Comprehensive imaging and Raman spectroscopy of carbonate globules from

130 P.B. Niles et al. / Earth and Planetary Science Letters 286 (2009) 122–130

Martian meteorite ALH 84001 and a terrestrial analogue from Svalbard. Meteorit.Planet. Sci. 42, 1549–1566.

Stiller, M., Rounick, J.S., Shasha, S., 1985. Extreme carbon-isotope enrichments inevaporating brines. Nature 316, 434–435.

Treiman, A.H., 1995. A petrographic history of Martian meteorite ALH84001 — 2 shocksand an ancient age. Meteoritics 30, 294–302.

Treiman, A.H., Romanek, C.S., 1998. Bulk and stable isotopic compositions of carbonateminerals in Martian meteorite Allan Hills 84001: no proof of high formationtemperature. Meteorit. Planet. Sci. 33, 737–742.

Valley, J.W., Eiler, J.M., Graham, C.M., Gibson, E.K., Romanek, C.S., Stolper, E.M., 1997.Low-temperature carbonate concretions in the Martian meteorite ALH84001:evidence from stable isotopes and mineralogy. Science 275, 1633–1638.

Warren, P.H.,1998. Petrologic evidence for low-temperature, possiblyfloodevaporitic originofcarbonates in the ALH84001 meteorite. J. Geophys. Res.-Planets 103, 16759–16773.

Yadav, D.N., 1997. Oxygen isotope study of evaporating brines in Sambhar Lake,Rajasthan (India). Chem. Geol. 138, 109–118.

Zheng, Y.F., 1990. Carbon–oxygen isotopic covariation in hydrothermal calcite duringdegassing of CO2 — a quantitative-evaluation and application to the Kushikino goldmining area in Japan. Miner. Deposita 25, 246–250.

Zolotov, M.Y., Shock, E.L., 2005. Formation of jarosite-bearing deposits through aqueousoxidation of pyrite at Meridiani Planum. Mars. Geophys. Res. Lett. 32.